|本期目录/Table of Contents|

[1]单玲玉1,2,闵 锋*1,等.全局阈值与局部阈值相结合的视网膜血管分割方法[J].武汉工程大学学报,2015,37(03):62-67.[doi:10. 3969/j. issn. 1674—2869. 2015. 03. 013]
 ,,et al.Segmentation method of retinal blood vessels via global threshold  and local threshold[J].Journal of Wuhan Institute of Technology,2015,37(03):62-67.[doi:10. 3969/j. issn. 1674—2869. 2015. 03. 013]
点击复制

全局阈值与局部阈值相结合的视网膜血管分割方法(/HTML)
分享到:

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
37
期数:
2015年03期
页码:
62-67
栏目:
机电与信息工程
出版日期:
2015-04-23

文章信息/Info

Title:
Segmentation method of retinal blood vessels via global threshold  and local threshold
文章编号:
1674—2869(2015)03—0062—06
作者:
单玲玉1闵 锋*1李延达3
1.武汉工程大学计算机科学与工程学院,湖北 武汉430205;2.智能机器人湖北省重点实验室(武汉工程大学),湖北 武汉430205;3.新洲一中,湖北 武汉430400
Author(s):
SHAN Ling-yu1 MIN Feng1 LI Yan-da3
1.School of Computer Science and Engineering, Wuhan Institute of Technology, Wuhan 430205, China;2.Hubei Key Laboratory of Intelligent Robot(Wuhan Institute of Technology), Wuhan 430205, China;3.Xinzhou NO.1 High school,Wuhan 430400,China
关键词:
视网膜血管全局阈值法局部阈值法匹配滤波
Keywords:
retinal blood vessel global threshold local threshold matched filter
分类号:
TP391.41
DOI:
10. 3969/j. issn. 1674—2869. 2015. 03. 013
文献标志码:
A
摘要:
针对视网膜血管图像使用单一阈值方法无法有效提取整体血管网络的问题,提出了一种全局阈值与局部阈值相结合的视网膜血管分割方法.该方法采用匹配滤波方法来增强图像,然后在增强图像上分别使用全局阈值二维最大熵与局部阈值移动平均算法对图像进行分割以获取视网膜的主血管与细小血管部分,最后通过区域连通性的判断,将视网膜的主血管与细小血管进行结合,分割出最终的血管网络.在Drive公共数据库上进行实验,结果表明,该方法分割得到的血管准确率达到93.56%,真阳性率值达到80.46%,均高于Perez算法,并且在细小部分更为丰富,能够得到较好的血管分割结果.
Abstract:
Aimed at the issue that the whole network of blood vessels can not be effectively extracted by a single threshold method, a novel segmentation method of retinal blood vessels image via global threshold and local threshold was proposed. Firstly, matched filter was used to enhance the retinal image. Then the main vessels and small vessels were respectively segmented through the global threshold of two-dimensional maximum entropy and the local threshold of moving average. Finally, the segmentation results of blood vessels were obtained via analyzing the region connectivity and combining the main vessels with small vessels. The accuracy of the segmentation method tested on Drive public database is 93.56%, and the value of true positive rate reaches 80.46%, both of which are higher than those by the Perez algorithm. In addition, the part of small vessels is more abundant, getting a better segmentation result of blood vessels.

参考文献/References:

[1] 李敏,罗洪艳,郑小林,等.一种改进的最大类间方差图像分割法[J].南京理工大学学报,2012,36(2):332-337.LI Min, LUO Hong-yan, ZHENG Xiao-lin, et al. Image segmentation based on improved otsu algorithm[J]. Journal of Nanjing University of Science and Technology, 2012,36(2):332-337.(in Chinese)[2] 张新明,张爱丽,郑延斌,等.改进的最大熵阈值分割及其快速实现[J].计算机科学,2011,38(8):279-283.ZHANG Xin-ming, ZHANG Ai?鄄li, ZHENG Yan?鄄bin, et al. Improved two?鄄dimensional maximum entropy image thresholding and its fast recursive realization[J]. Computer Science, 2011,38(8):279-283.(in Chinese)[3] 姚畅,陈后金,荆涛,等.一种基于改进的PCNN的视网膜血管树提取方法[J].光电子.激光,2011,22(11):1745-1750.YAO Chang, CHEN Hou-jin, JING Tao, et al. Extraction of blood vessel tree in retinal image based on improved PCNN[J]. Journal of Optoelectronics Laser, 2011,22(11):1745-1750.(in Chinese)[4] 黄琳,沈建新,罗煦.视网膜图像中的血管自适应提取[J].中国制造业信息化,2009,38(1):64-67.HUANG Lin, SHEN Jian-xin, LUO Xun. The Automatic extracting blood vessels in retinal image[J]. Manufacture Information Engineering of China,2009,38(1):64-67.(in Chinese)[5] 许立腾,徐向民.基于二维最大熵阈值分割的钙化点检测算法[J].计算机仿真,2010,27(9):255-257.XU Li-teng, XU Xiang-min. A calcification detection method based on two-dimensional entropic thresholding[J]. Computer Simulation,2010,27(9):255-257.(in Chinese)[6] Martinez-Perez M E, Hughes A D, Thom S A, et al. Segmentation of blood vessels from red-free and fluorescein retinal images[J]. Medical Image Analysis, 2007, 11(1):47-61.[7] 张东波,尚星宇.病变视网膜图像的血管骨架提取方法研究[J].电子测量与仪器学报,2011,25(9):749-755.ZHANG Dong-bo, SHANG Xing-yu. Extracting blood centerline adapted for retinal fundus images with pathologies[J]. Journal of Electronic Measurement and Instrument, 2011, 25(9): 749-755.(in Chinese)[8] VIERGEVER M, LUIJTEN P. DRIVE: Digital retinal images for vessel extraction[EB/OL]. http://www.isi.uu.nl/Research/Database/DRIVE/. [2010-09-01].

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2015-01-17基金项目:湖北省自然科学基金资助项目(2012FFA099;2014CFA130);武汉工程大学科学研究基金(K201401)作者简介:单玲玉(1989-),女,河南新乡人,硕士研究生.研究方向:图像处理.*通信联系人.
更新日期/Last Update: 2015-04-27